MixGPT2_10K_fromB_BFall_10KGen_topP_0.75_noaddedB
This model is a fine-tuned version of bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0624
- Accuracy: 0.9941
- F1: 0.9342
- Precision: 0.9971
- Recall: 0.8788
- Roc Auc Score: 0.9393
- Tpr At Fpr 0.01: 0.8718
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall | Roc Auc Score | Tpr At Fpr 0.01 |
|---|---|---|---|---|---|---|---|---|---|
| 0.0103 | 1.0 | 6875 | 0.0219 | 0.9942 | 0.9359 | 0.9807 | 0.895 | 0.9471 | 0.7034 |
| 0.0064 | 2.0 | 13750 | 0.0368 | 0.9942 | 0.9359 | 0.9922 | 0.8856 | 0.9426 | 0.8102 |
| 0.0019 | 3.0 | 20625 | 0.0487 | 0.9942 | 0.9355 | 0.9977 | 0.8806 | 0.9403 | 0.88 |
| 0.0005 | 4.0 | 27500 | 0.0574 | 0.9942 | 0.9352 | 0.9944 | 0.8826 | 0.9412 | 0.8494 |
| 0.0 | 5.0 | 34375 | 0.0624 | 0.9941 | 0.9342 | 0.9971 | 0.8788 | 0.9393 | 0.8718 |
Framework versions
- Transformers 4.29.1
- Pytorch 1.9.0+cu111
- Datasets 2.10.1
- Tokenizers 0.13.2